The work involved functional MRIs of jazz musicians while they were playing both scales and a blues piece and while they were improvising around each. The idea is that the difference between the two provides insight into what the brain is doing during creativity. Some of the results were unsurprising - the medial prefrontal cortex (which has to do with one's sense of self) became more active while improvising and the lateral prefrontal cortex (which has to do with self-censorship) became less active. But all of the sensory-motor areas of the brain became more active, including regions one would think were completely unrelated, e.g. the visual regions. And the limbic areas (which have to do with emotion) became less active, suggesting improvisation is a dispassionate activity.
He also suggested that music is more complex than language, which I think may be the answer to my pondering the other night about why it takes me more concentration to listen to orchestral music than it does to things with words. I haven't thought about how listening to vocal music in languages I can't understand (something I actually do quite a lot of) fits in.
Limb did an excellent job of presenting technical material to amateurs. That is, one didn't need to really know much of anything about either neurology or music to understand his lecture. A good speaker and a subject I find inherently fascinating always makes for an interesting hour.
By the way, the LoC is recording the lectures in their "Music and the Brain" series and they are going to have webcasts out a couple of weeks after each. The series goes on for two years, so there should be a real wealth of material by the time it's done.